Bill Gates is sounding one of his starkest warnings yet about artificial intelligence, saying the technology could become an extraordinarily dangerous weapon if increasingly powerful AI systems fall into the hands of people intent on causing mass harm.
The Microsoft co-founder said AI is already powerful enough to potentially help trigger events resulting in as many as one billion deaths — and argued that governments cannot leave the technology industry to police itself.
“AI is certainly powerful enough to drive events that cause a billion deaths,” Gates said during an interview with NBC’s “Meet the Press” that aired Sunday, September 27.
His warning came after host Kristen Welker asked whether AI could become powerful enough to wipe out humanity.
Gates stopped short of predicting human extinction.
Instead, he focused on a scenario in which people deliberately misuse advanced AI systems, saying the combination of sophisticated technology and malicious actors presents a threat unlike previous weapons.
“There’s never been a weapon as powerful as the combination of people with ill intent using the latest AI tools,” Gates said.
Gates Says Tech Companies Cannot Police Themselves
Gates also rejected the idea that AI companies should be solely responsible for deciding how their increasingly powerful systems are controlled.
“No one thinks self-regulation is enough,” he said.
Gates argued that elected officials and law enforcement agencies need to become directly involved in determining what safeguards, monitoring requirements and rules should govern advanced AI.
“You need law enforcement and the politicians to get into the discussion about what safeguards and monitoring look like,” Gates said, adding that some protections would have to become mandatory rather than voluntary.
His concern centers partly on how difficult it could become to distinguish beneficial scientific work from research intended to cause harm.
Gates pointed to medicine and biotechnology as an especially troubling example.
“I wish it was easier to distinguish good intent, like I want to design a new drug for a disease, versus negative intent, which is I want to design a tool of bioterrorism,” he said.
The problem, according to Gates, is that the technical work involved in developing lifesaving drugs could overlap with the knowledge needed to create dangerous biological agents.
Because those activities can appear similar, Gates said sophisticated AI systems may eventually require extensive monitoring of how they are being used.
His comments reflect a growing concern among researchers that AI could dramatically lower the level of expertise needed to perform tasks that once required years of specialized training.
That does not mean current consumer chatbots can simply create a biological weapon on command. Rather, researchers and policymakers are debating whether increasingly advanced future models could accelerate scientific work enough to make dangerous capabilities easier to access.
Gates Has Warned About Bioterrorism Before
The billionaire philanthropist has repeatedly raised concerns about biological threats.
Earlier this year, Gates warned that AI could make it easier for a relatively small group to develop a dangerous biological weapon.
His latest comments go further by putting a staggering potential death toll on what he considers a worst-case misuse scenario.
But Gates is not calling for AI development to end.
In fact, he remains one of the technology’s most outspoken supporters when it comes to health care, agriculture, education and economic development.
Earlier in September, Gates told Reuters that AI has enormous potential to improve lives but argued governments around the world are poorly prepared for the disruption the technology could bring.
He cited job losses, addictive AI products and malicious attacks among the risks governments should begin addressing.
At the Gates Foundation’s annual Goalkeepers event in New York this month, Gates also promoted what he called the “smart use of AI” to improve health care, education and farming in poorer countries.
The foundation has committed $1 billion toward AI-related initiatives and is working with companies including Google, Anthropic and OpenAI on projects aimed at making AI systems more useful in languages that have historically been underrepresented online.
In other words, Gates’ position is not that AI itself should disappear.
His argument is that the potential benefits are enormous — but so are the consequences if safeguards fail.
Trump Takes a Different Approach
The warning comes amid an increasingly heated national debate over whether Washington should regulate the technology more aggressively or prioritize rapid development as the United States competes with China.
President Donald Trump has emphasized maintaining American leadership in AI and has pushed back against some of the most dire warnings surrounding the technology.
Speaking earlier this month, Trump described some AI fears as exaggerated while stressing that the United States cannot afford to lose its technological advantage over China.
During his September 22 address to the United Nations, Trump said the United States rejected attempts to create what he described as a global system of control over AI.
His administration has also emphasized innovation and technological leadership, although Trump said the Justice Department could intervene if AI companies create serious problems requiring government action.
The White House has repeatedly described the competition with China as an AI race with major economic, military and geopolitical consequences.
A Trump administration technology report earlier this year said the country that dominates AI could shape industries, international standards and strategic advantages for decades.
Other Political Leaders Raise Concerns
Former President Barack Obama has taken a more cautious position.
Reuters reported earlier this month that Obama warned during a private gathering that rapidly developing AI systems could become dangerous without greater oversight.
Obama reportedly urged Democrats to begin developing a broader policy framework for the technology and encouraged House Minority Leader Hakeem Jeffries to promote a public debate over AI safety and its effect on the economy.
The issue does not divide Americans neatly along party lines, however.
A Reuters/Ipsos poll published September 22 found widespread concern about whether technology companies can adequately police themselves.
Seventy-three percent of Americans surveyed said they were concerned AI companies were not doing enough to prevent potentially catastrophic consequences.
The same poll found 55% supported slowing the development of advanced AI systems, while 73% said safe and responsible development should be prioritized over winning the international AI race.
The online survey included 1,277 U.S. adults and carried a margin of error of about three percentage points.
The Bigger Question Facing Washington
Gates’ comments highlight the dilemma now facing lawmakers.
AI systems are already being used to accelerate medical research, write computer code, analyze huge amounts of information and automate work that previously took humans hours or days.
At the same time, researchers are studying how those same capabilities could be exploited for cyberattacks, fraud, disinformation and potentially biological threats.
That creates a difficult balancing act.
Regulate too aggressively, critics argue, and the United States could slow innovation while competitors such as China continue racing ahead.
Move too slowly, Gates and other advocates of stronger safeguards warn, and governments may discover they have lost control of the technology only after something goes badly wrong.
Gates believes the speed of development means governments need to begin answering those questions now.
His billion-death warning remains a hypothetical worst-case scenario — not a prediction that such a catastrophe is inevitable.
But his central message was unmistakable: as AI becomes more powerful, he believes relying on technology companies to voluntarily control its most dangerous uses will no longer be enough.
No different than your bio engineered mosquitoes.